ADSORPTION OF MOLYBDATE ON NANO-BALL ALLOPHANE

Clay Science 11,189-204 (2000)

ADSORPTION OF MOLYBDATE ON NANO-BALL ALLOPHANE

ELSADIG AGABNA ELHADI, NAOTO MATSUE and TERUO HENMI

Faculty of Agriculture, Ehime University, 3-5-7 Tarumi, Matsuyama 790-8566, Japan

(Received February 7, 2000, Accepted June 21, 2000)

ABSTRACT

Adsorption behavior of Mo (Na2 MoO4) on three nano-ball shaped allophanesamples was different between lower (up to 0.1mM) and higher (0.2 to 1.6mM) Moconcentrations. The difference may be due to chemical species of Mo in solution: atlower concentration, Mo exists as monomeric H2MoO4, HMoO4- and MoO42-, andat higher concentrations as polymeric forms. The monomeric molybdate species,especially the anionic species, had very strong affinity for the allophane at lower pHwhere the allophane had much surface positive charge and essentially no negativecharge. All of the molybdate added at equilibrium pH 3.7 to 3.9 and initialconcentration up to 0.1mM, 200 mmol kg-1, was adsorbed on allophane sampleswith lower Si/Al ratio. With increasing pH, the affinity or amount of adsorptiondecreased markedly. At equilibrium pH near 5, adsorption isotherm was Langmuirtype, whereas at higher pH such as 6 or 8 the isotherm was linear or Freundlichtype. The strong interaction between molybdate anion and positively chargedallophane was concluded as electrostatic one followed by ligand exchange reactionbetween Mo-0- and aluminol groups. At higher Mo concentrations, more than0.2 mM, shape of the isotherms were linear indicating further polymerization re-action on the allophane surfaces, and partial destruction of allophane structure wasobserved as Si and Al release with the adsorption.

Key words: Adsorption, Molybdate, Nano-ball allophane, Ligand exchange

INTRODUCTION

Molybdenum is an essential micronutrient for plants and animals, but excessive amountcan have deleterious effects on various organisms. Concentration of Mo in the environmentis significantly enhanced by anthropogenic input from coal resource development fly ash,sewage sludge, and hard rock mining activity. The amount of Mo adsorbed by soils isdependent on Mo concentration, pH and the mineralogical composition of the soils.Molybdenum occurs in aerated soil as molybdate (MoO42-) or its protonated forms

(HMoO4- and H2MoO4) or as Mo(OH6), depending on pH (Lindsay, 1979; Gupta andLipsett, 1981). The pKl and pK2 values of H2MoO4 are 4.00 and 4.24, respectively,therefore above pH 5, Mo exists mainly as Mo042- (Lindsay, 1979).

Reactions of molybdate have been described as some of the most complex ones (Cottonand Wilkinson, 1980). It forms many complexes and polymerized species at high con-centrations, but in solution of less than 0.1 mM the polymeric forms are not of greatimportance (Jenkins and Wain, 1963). Phelan and Mattigod (1984) reported that no

190 E.A. Elhadi et al.

polymeric form existed up to 11 ppm (about 0.1 mM). Carpeni (1947) concluded that onlymonomeric molybdate ions exist in solution below 0.3 mM at all pH values. However, no

general agreement exists with respect to this postulation of Carpeni (1947) or with respectto the condensation products of the monomer. It is believed that the heptameric anionMo7O246-is the first product and that further condensation produces the octameric anionMo8O264-(Cotton and Wilkinson, 1980). Dimolybdate anion is discarded as product sinceit is only stable in non-aqueous solvents (Day et al., 1977).

The mechanism of molybdate adsorption on Al and Fe oxides was suggested to beligand exchange with surface hydroxyl groups (Ferreiro et al., 1985). According to thismechanism, molybdate become specifically adsorbed as inner surface complexes, which bydefinition contain no water molecules between the adsorbing molybdate and the surfacefunctional groups (Sposito, 1984). Molybdate adsorption has been investigated for avariety of Al and Fe oxides, clay minerals and soils, but relatively few studies have beenconducted on allophanic or volcanic ash soil, and actually only a few studies have beenappeared in literatures (Theng, 1971; Gonzalez et al., 1974). However, these studies wereconducted using soil, which contain components other than allophane.

Allophane is a main component of volcanic soils, and is a series name of naturallyoccurring amorphous hydrous aluminosilicate with various chemical compositions (Wadaand Harward, 1974). The aluminosilicate had not been believed to have definite mor-

phology and structure. However, Henmi and Wada (1976) found that allophane at leastseparated from volcanic ash soils and weathered pumice grains had definite hollowspherical morphology (nano-ball with diameter of about 5 nm) with narrow range ofchemical compositions from 1:2 to 2:2 with respect to Si/Al atomic ratio. Thereafter,allophane with hollow spherical morphology is referred to as nano-ball allophane. Detailedmineralogical analysis and physicochemical properties have been carried out for carefullycollected pure nano-ball allophane, so as to make clear the chemical structure of wall ofthe ball in relation to its Si/Al ratio (Henmi, 1980; Henmi et al., 1983; Henmi and Huang,1985). The wall structure of nano-ball allophane has been proposed as aluminum-neso-silicate structure composed of curved gibbsite sheet with monomeric SiO4 tetrahedraattached to it (Parfitt and Henmi, 1980). Study using NMR has proved the proposedchemical structure of nano-ball allophane (Shimizu et al., 1988). Henmi et al. (1997)showed that the chemical structure of nano-ball allophane with Si/Al ratio of 1:2 isfundamental structure and allophane with Si/Al higher than 1: 2 has excess Si atom weaklyattached to the SiO4 tetrahedra in the structure.

The purpose of this paper is to study the adsorption behavior and mechanism ofmolybdate on nano-ball allophane, using the recent information about the detailedmorphology and chemical structure of the wall of the aluminum silicate.

MATERIALS AND METHODS

Sample Preparations

Allophane samples were separated from inner part of pumice grains collected from

different places in Japan. The sample KyP was collected from Kurayoshi, Tottori pre-

fecture, KiP from Kitakami, Iwate prefecture, and KnP was collected from Kakino,

Molybdate Adsorption on Allophane 191

Kumamoto prefecture. Fine clay fraction (<0.2 ƒÊm) was separated from the inner part of

the pumice grains after removing the outer part in order to eliminate any possible

contamination of the sample with impurities such as volcanic glass, opaline silica and

imogolite. The separation was carried out by ultrasonification at 28 kHz and then dis-

persion at pH 4 for sample with low Si/A1 ratio (KyP and KiP) and at pH 10 for KnP

sample with high Si/Al ratio (Henmi and Wada, 1976). The collected samples were

flocculated using NaCl, washed with water to remove excess salt and then subjected to

freeze drying. Collected freeze dry samples were analyzed using X-ray dffraction (XRD),

infrared spectroscopy (IR) and thermal analysis; the results obtained suggest that these

samples do not contain impurities described above. The Si/Al atomic ratio was 1.34:2,

1.42:2 and 1.98 :2 for KyP, KiP and KnP, respectively. Schematic chemical structures for

unit particles of nano-ball allophane with Si/Al atomic ratio of 1:2 and that with more

than 1:2 are presented in Fig. 1.

Molybdate Adsorption ExperimentMolybdate adsorption behavior was studied by equilibrating 50mg freeze-dried

allophane sample with 100mL of mixed solution of Na2MoO4, NaC1 and water. Theinitial molybdate concentration of the mixed solution was varied from 0 to 1.6 mM, whiletotal Na concentration was kept at 10mM by adjusting the amount of NaCl added. ThepH of the mixed solution was adjusted by using either HC1 or NaOH to bring the pH inthe range from 3 to 10. The mixture was shaken for 24 h, and then centrifuged at 5000rpm for 20 min. The pH of the supernatant was measured, and the concentration ofMo, Si and Al were determined with atomic absorption spectroscopy. The amount ofmolybdate adsorbed was calculated from difference between the initial and final (equi-librium) Mo concentrations of the supernatant. Molybdate adsorption data was analyzedaccording to Langmuir and Freundlich equations. Langmuir equation is shown below,

X =XmKC/(1+KC)

where X=amount of molybdate adsorbed (mmol kg-1), K=constant related to bindingenergy (L mmol-1), Xm=maximum adsorption (mmol kg-1) and C=equilibriummolybdate concentration (mmol L-1). The adsorption data was plotted according to thelinear form of Langmuir equation.

C/X=1/XmK+C/Xm

Freundlich equation is,

X=kC l/n

where X is molybdate adsorbed per unit mass (mmol kg-1), C is the equilibriummolybdate concentration (mmol L-1), whereas k and n are empirical constants. Theadsorption data was plotted according to the logarithm form,

log X=log k+1/n log C.

192 E.A. Elhadi et al.

B

A

C

Si/Al=1:2

Si/Al>1:2

FIG. 1. Morphology and chemical structure of nano-ball shaped allophane (A: molecular morphologyin section; B: atomic arrangement near the pore of hollow particles; C: atomic arrangement in thecross section of allophane particles).


RESULTS AND DISCUSSION

Molybdate Adsorption IsothermAdsorption isotherms for the three nano-ball allophane samples at initial molybdate

concentrations from 0.2 to 1.6 mM are shown in Fig. 2. For the three allophane samples,amount of molybdate adsorption decreased with increasing pH, and at initial solution pHabove 8, the adsorption became essentially zero over the concentration range. This is inagreement with the result reported by McKenzie (1983) on the adsorption of molybdate on

goethite. With increasing pH, negative charges (Si-O-) develop on the allophane samples,and electrostatic repulsion force might prevent the approach of molybdate anion to thenano-ball allophane surfaces. For KyP and KiP samples with lower Si/Al ratio, at initial

pH of 3 and 4, isotherms seemed to have positive intercepts, whereas the positive interceptwas observed only at initial pH 3 for KnP sample with higher the ratio. The positiveintercept indicates strong interaction between molybdate and allophane at very lowmolybdate concentration region.

Langmuir and Freundlich plots for the isotherms are shown in Fig. 3a and 3b, re-spectively. For the Langmuir case, shape of the plots were curve rather than straight line,indicating that the molybdate adsorption on allophane is not monolayer type, and not a1:1 type between adsorbate and adsorption site. Freundlich plots were fitted to straightline better than the Langmuir plots, and obtained Freundlich parameters are given inTable 1. Sabine Goldberg and Foster (1998) reported that Mo adsorption data fitted toboth Langmuir and Freundlich equation, with better ffitness to Freundlich equation.Parameter n is related to the strength of the adsorption, while k value is related to bothstrength and amount of the adsorption. Table 1 shows that the values of n and k decreasedwith increasing initial solution pH. On the other hand, the trend between the allophanesamples with regard to the n and k values was in the order, KyP>KiP>KnP, with thelowest values of n and k for KnP under the three pHs. This order is ascribed to highercontent of aluminol groups, Al(OH)(OH2), in KyP and KiP samples, because the amountof aluminol groups per unit mass decreases with increasing Si/Al ratio of the allophanesamples (Son et al., 1998). Results obtained so far for adsorption of phosphate (Johan etal., 1997) and borate (Son et al., 1998) on nano-ball allophane support the fact thataluminol groups, Al(OH)(OH2), are responsible for adsorption of the anions. As men-tioned in INTRODUCTION, Mo exists as polymerized forms in this range of molybdateconcentrations, so the better fitness to Freundlich equation may indicate further poly-merization of Mo species on the allophane surfaces.

Molybdate adsorption experiment at lower initial molybdate concentration (up to 0.1mM) was carried out to know in more detail the initial strong interaction indicated in Fig.2. Initial molybdate concentrations were 0, 0.02, 0.04, 0.06, 0.08 and 0.1 mM, and all ofthe molybdate may present as monomeric forms at these concentrations (Phelan andMattigod, 1984). Isotherms presented in Fig. 4 clearly show that the amount of molybdateadsorption increased with decreasing pH of the initial molybdate solution for the threesamples. At initial pH 3, almost all the molybdate added were adsorbed on KyP and KiPsamples with lower Si/Al ratios, and for KnP with higher the ratio more than 70% wasadsorbed. Here, initial molybdate concentration of 0.1 mM corresponds to 200 mmol ofmolybdate per one kg of allophane sample, and equilibrium solution pH of the blank run

194 E. A. Elhadi et al.

KyP

KiP

KnP

Equilibrium concentration (tmol L-1)

FIG. 2. Molybdenum adsorption isotherms on allophane samples at higher Mo

concentrations.


KyP

Ki P

KnP

C (ƒÊmo1 L-1)

FIG. 3a. Langmuir plots for the adsorption of Mo on allophane samples at pH 3

and 4.

for the initial pH 3 were 3.7, 3.9 and 4.2 for KiP, KyP and KnP samples, respectively. At

these pH values, KiP, KyP and KnP had positive charges of 400, 400 and 150 mmol kg-1,

respectively. These observations indicate that at initial solution pH of 3, molybdate were

very strongly or specifically adsorbed on positive charge site of nano-ball allophane,


KyP

KiP

KnP

Log C

FIG. 3b. Freundlich plots for the adsorption of Mo on allophane samples at

different pH.

Al-OH2+, which locate at pore region of the wall (Fig. 1). Under the condition, initial pHof 3, amounts of positive charge on KiP and KyP samples were twice of the molybdateadded, so all the molybdate added could be adsorbed on the samples even as divalentanion, MoO4 2-. Detailed adsorption mechanism will be discussed in the following section.


TABLE 1. Freundlich parameters for Mo adsorption on nano-ball

allophane samples

pH 3 pH 4 pH 6

On the other hand, for KnP at the same initial pH, some of the molybdate remained insolution phase even at initial molybdate concentration of 0.02 mM, which corresponds to40 mmol of Mo per one kg of allophane sample. This means that the adsorption ofmolybdate on positive charge site on KnP with higher Si/Al ratio is weaker than KyP andKiP. However, molybdate still has much greater affinity for KnP than Cl-has, becausemore than 70% added were adsorbed under 10 mM NaCl background solution (Fig. 2).

Among the isotherms shown in Fig. 4, that for KnP at initial pH 3 and those for KyPand KiP at pH 4 were fitted to Langmuir equation, and the others to Freundlich; those atinitial pH 3 for KyP and KiP were impossible to analyze because equilibrium molybdateconcentrations were essentially zero. The calculated Langmuir Xm values for the threecases were roughly the same with the corresponding amount of positive charge on eachsample. The Xm and amount of positive charge in mmol kg-1 were 250 and 220 for KnPat initial pH 3, 166 and 180 for KyP at initial pH 4, and 200 and 180 for KiP at initial pH4. This means that when nano-ball allophane has positive charge, strong interaction existsbetween Al-OH2+of allophane and molybdate anion. For KyP and KiP at initial pH 3,the interaction is considered to be very strong, because no negative charge exists onallophane at these conditions. Initial reaction force between negatively charged molybdateand positively charged aluminol group is electrostatic, but after adsorption, further reactionmay occur. At higher pH region where isotherms fitted to Freundlich equation andallophane samples have small or essentially no positive charge but have some negativecharge, the interaction is assumed to be much weaker and not to be an electrostatic one.

Adsorption MechanismFigure 5 shows the change in solution pH after molybdate adsorption at initial

molybdate concentrations of 0.2 to 1.6 mM. The pH value for the blank run (zeromolybdate concentration) was considered as the initial pH for molybdate and allophanebefore adsorption. At initial pH 3, KyP and KiP samples showed decrease in pH afteradsorption. However, no apparent change at initial pH 4 was observed with the ad-sorption. On the other hand, decrease in pH for KnP was observed at initial pH 3 and pH4. Positive change occurred at initial pH 6 for the three samples, but the change was verysmall for KnP sample. Negative change in pH after adsorption may be attributed torelease of hydrogen ion during adsorption while the positive change at pH 6 may indicaterelease of hydroxyl ion during the adsorption. At this initial molybdate concentrationrange, release of Si and Al with molybdate adsorption was observed at initial pH of 3 and


KyP

KiP

KnP

Equilibrium concentration (ƒÊmol L-1)

FIG. 4. Molybdenum adsorption isotherms on allophane samples at lower Mo

concentrations.


KyP

KiP

KnP

Na2MoO4 concentration (mM)

FIG. 5. Change in pH after Mo adsorption.


Kyp

KiP

KnP

Mo adsorbed (umol g1)

FIG. 6a. Silicon released from allophane samples with Mo adsorption .


KyP

KnP

KnP

FIG. 6b. Aluminum released from allophane samples with Mo adsorption.


4. As shown in Figs. 6a and 6b, the amount of Si and Al released increased almost linearlywith increase in the amount of molybdate adsorbed. Johan et al (1997) reported that therewas a linear relationship between P adsorbed and Si released, and suggested replacementreaction between phosphate and polymeric Si.

The ratios of Al released to molybdate adsorbed were calculated as 0.52, 0.59 and 0.57at pH 3, and 0.96, 0.39 and 0.52 at pH 4 for KyP, KiP and KnP, respectively. The releaseof Al may be attributed to partial destruction of the allophane at the pore region of nano-ball wall after molybdate adsorption. This is in agreement with the result obtained by Sonet al (1998) on the adsorption of boron on allophane at initial pH 4. On the other hand,the ratios of Si released to Mo adsorbed are 0.15, 0.12 and 0.24 at pH 3 and 0.29, 0.11and 0.11 at pH 4 for KyP, KiP and KnP, respectively. Release of Si and Al at pH 6 wasconstant and very small irrespective of adsorption. The observed pH change and release ofSi and Al together with the linear adsorption isotherm (Fig. 2) indicate that, at highermolybdate concentrations, complicated reaction including adsorption, polymerization ofmolybdate on allophane surface and destruction of allophane structure occurred between

polymeric molybdate and nano-ball allophane.For the lower initial molybdate concentration region (up to 0.1 mM) where all of the

molybdate existed as monomeric form, however, change in solution pH with molybdateadsorption was not observed for all pHs and samples. In addition, release of Si and Alduring the adsorption was not observed for all the cases. These facts together with theisotherm in Fig. 4 indicate that the monomeric molybdate adsorption on allophane atlower pH and lower concentration region is ligand exchange reaction between aluminol

groups and Mo-0- group of monomeric molybdate. Previous studies (Rayes and Jurinak,1967: Theng, 1971; Gonzalez et al., 1974) indicate that only monomeric species will beadsorbed. Zhang and Sparks (1989) found that molybdate anion react primarily with the

protonated site, Al-OH2+, and not with the neutral one. While according to the model ofSpanos and Lycourghiots (1995) deposition of Mo species on y-alumina take place mainlyvia adsorption of Mo04 2- on sites created by the protonated surface hydroxyls. The linearisotherms with negative intercepts for initial pHs of 6 and 8 (Fig. 4) may indicate somedeposition or polymerization reaction of molybdate on allophane surfaces.

At the pore region of nano-ball allophane, Al-OH, Al-OH2 and Al-OH2+exist atlower pH (Fig. 1), and the ligand exchange reaction between positively charged aluminol

group and Mo-O- group at lower pHs is schemed as:

Al-OH2++-O-Mo→ Al-O-Mo+ H2O (1)

The reaction releases H2O and causes no change in solution pH. However, the observed

constancy in pH with the molybdate adsorption does not necessarily mean that no H+or

OH- was released with the adsorption. Because molybdate exist as both neutral and

anionic species at pH around 4, the selective adsorption of anion species causes shift in

chemical equilibrium between neutral and anionic molybdate species to release H+in

solution. Therefore, other ligand exchange reactions to release OH- may occur simul-

taneously:

Al-OH+-O-Mo→ Al-O-Mo+ OH- (2)


Al-OH2+-O-Mo→ [Al-O-Mo]-+ H2O (3)

reaction in equation (3) will produces OH- because part of the negatively chargedadsorption complex then may be protonated. The 100% adsorption for KiP and KyP atinitial pH 3 indicates that equation (1), reaction between positive and negative functionalgroups, is the main reaction. If the molybdate anion adsorbed on allophane with bidentateform, reactions in the equation (2) or (3) also took place in addition to reaction inequation (1). Scheme for the bidentate ligand exchange reaction including reactions (1) and

(2), for example, is written as below.

(4)

The O-O distance in the molybdate molecule (0.288nm) is similar to that betweenaluminol groups at pore region of allophane, therefore the bidentate adsorption is possible.

REFERENCES

CARPENI, G.(1947) Sur la constitution des solution aqueuses d'acid molybdique et de molybdatesalclins. I. V. Conclusions general. Bull. Soc. Chim., 14, 501-503.

COTTON, F. A. and G. WILKINSON (1980) Advanced Inorganic chemistry. Wiley and Sons, New York;p.855.

DAY, V. W., M. F. FREDRICH, W. G. KLEMPERER and W. SHUM (1977) Synthesis and characterization ofthe dimolybdate ion. J. Amer. Chem. Soc., 99, 6146-6148.

FERREIRO, E. A., A. K. HELMY and S. G. dE BUSSETTI (1985) Molybdate sorption by oxides of aluminiumand iron. Z. Pfianzenernaehr. Bodenk., 148, 559-566.

GONZALEZ, B. R., H. APPLET, E. B. SCHALSCH and F. T. BINGHAM (1974) Molybdate adsorptioncharacteristics of volcanic-ash-derived soils in Chile. Soil Sci. Soc. Am. Proc., 38, 903-906.

GUPTA, U. C. and J. LIPSETT (1981) Molybdenum in soils, plant and animals. Adv. Agron. 34, 73-115.HENMI, T.(1980) Effect of SiO2/Al2O3 ratio on thermal reactions of allophane. Clays Clay Miner., 28,

92-96.HENMI, T. and P. M. HUANG (1985) Removal of phosphorus by poorly ordered clays as influenced by

heating and grinding. Applied Clay Sci., 1, 133-144.HENMI, T., N. MATSUE and E. JOHAN (1997) Change in the surface acidity of allophane with low Si/Al

ratio by the reaction with ortho-silicic acid. Jpn. J. Soil. Sci. Plant Nutr., 68, 514-520 (inJapanese).

HENMI, T. and K. WADA (1976) Morphology and Composition of allophane. Am. Miner., 61, 379-390.JENKINS, I. L. and A. G. WAIN (1963) Molybdenum species in acid solution. J. Appl. Chem., 13, 561-

564.JOHAN, E., N. MATSUE and T. HENMI (1997) Phosphate adsorption on nano-ball allophane and its

molecular orbital analysis. Clay Science, 10, 259-270.LINDSAY, W. L.(1979) Chemical equilibria in soils. Wiley—Interscience, New York.MCKENZIE, R. M.(1983) The adsorption of molybdenum on oxide surfaces. Aust. J. Soil Res., 21,

505-513.PARFITT, R. L. and T. HENMI (1980) Structure of some allophane from New Zealand. Clays Clay

Miner., 28, 285-294.PHELAN, P. J. and S.V. MATTIGOD (1984) Adsorption of molybdate anions (MoO421-) by sodium-

saturated kaolinite. Clays Clay Miner., 32, 45-48.REYES, E. D. and J. J. JURINAK (1967) A mechanism of molybdate adsorption on a Fe203. Soil Sci. Soc.

Am. Proc., 31, 637-640.


SABINE GOLDBERG and H.S. FOSTER (1998) Factor affecting molybdenum adsorption by soil andminerals. Soil Science, 163, 109-114.

Smwzu, H., T. WATANABE, T. HENMI, A. MASUDA and A. SArro (1998) Studies on allophane andimogolite by high-resolution solid-state 295i- and 27 Al-NMR and ESR. Geochem. J., 22, 23-31.

SON, L. T., N. MATSUE and T. HENMI (1998) Boron adsorption on allophane with nano-ball mor-phology. Clay Science, 10, 315-225.

SPANOS, N. and LYCOURGruoTIS (1995) Co deposition of Mo(VI) species and Nit+ ions on the y-alumina surface: Mechanistic model. J. Colloid Interface Sci., 171, 306-318.

SPOSrTO, G. (1984) The surface chemistry of soil. Oxford Univ. Press. New York.THENG, B. K. G. (1971) Adsorption of molybdate by some crystalline and amorphous soil clays. New

Zealand J. Sci., 14, 1040-1056.WADA, K. and M.E. HARwARD (1974) Amorphous clay constituents in soils. Adv. Agron., 26, 211-

260.ZHANG, P. C. and D. L. SPARKS (1989) Kinetics and mechanism of molybdate adsorption/desorption

at the goethite/water interface using pressure jump relaxation. Soil Sci. Am. J., 35, 1028-1034.

Documents

ADSORPTION OF MOLYBDATE ON NANO-BALL ALLOPHANE